RecruitLingo

Help me evaluate an AI tool that screens, ranks or matches candidates, before we rely on it. THE TOOL: {{name and what the vendor says it does}} WHERE IT SITS IN OUR PROCESS: {{e.g. ranks inbound applicants, scores resumes against the JD, auto-rejects below a threshold, screens by video}} WHAT WE'D DO WITH ITS OUTPUT: {{is a human reviewing everything, or is it filtering people out before anyone looks}} WHERE WE HIRE: {{jurisdictions}} Give me: 1. The questions to put to the vendor: what the model was trained on, what it actually predicts, how it was tested for disparate impact, who audited it and when, and whether they'll share the results. 2. What could produce biased outcomes here even with protected characteristics removed — the proxies that carry the same signal. 3. What we should test ourselves before trusting it, and how to run that test with the data we have. 4. Our likely obligations, by jurisdiction: bias-audit requirements, candidate notification, the right to request human review, and record-keeping. Flag which of these change often and need checking with counsel. 5. What to tell candidates, and where that disclosure needs to appear. 6. The signs this tool is quietly narrowing the pipeline toward people who resemble past hires, and how we'd notice. 7. The line beyond which a human must make the decision, not the tool. This is a starting point for a conversation with legal and HR, not legal advice.